马丽,董唯光,梁金平,张晓东

• 1:兰州交通大学自动化与电气工程学院

摘要(Abstract)：

提出一种基于流形距离的局部线性嵌入算法,以流形距离测度数据间的相似度,选择各样本点的近邻域,解决了欧氏距离作为相似性度量时对邻域参数的敏感性.在MDLLE算法中引入最大边缘准则(maximum margin criterion,MMC)来构建最优平移缩放模型,使得算法在保持LLE局部几何结构的同时,具有MMC准则判别能力.通过正交化低维特征向量可消除降维过程中的噪声影响,进而提高算法的监督判别能力.由实验结果得到,所提出的方法具有良好的降维效果,能有效避免局部降维算法对邻域参数的敏感.随机投影独立于原始高维数据,将高维数据映射到一个行单位化的随机变换矩阵的低维空间中,维持映射与原始数据的紧密关系,从理论上分析证明了在流形学习算法中采用随机投影可以高概率保证在低维空间保持高维数据信息.

关键词(KeyWords)： 流形学习算法;邻域选择;流形距离;正交判别;局部线性嵌入;随机投影

Abstract:

Keywords:

基金项目(Foundation): 甘肃省科技支撑基金资助项目(1204GKCA038);; 甘肃省财政厅基本科研业务费资助项目(213063)

作者(Author): 马丽,董唯光,梁金平,张晓东

参考文献(References)：

• [1]ROWEIS S,SAUL L.Nonlinear dimensionality reduction by locally linear embedding[J].Science,2000,290(5500):2323—2326.
• [2]TENENBAUM J B,SILA A V,LANGFORD J C.A global geometric framework for nonlinear dimensionality reduction[J].Science,2000,290(5500):2319—2323.
• [3]孟德宇,徐宗本,戴明伟.一种新的有监督流形学习方法[J].计算机研究与发展,2007,44(12):2072—2077.
• [4]DONOHO D L,GRIMES C.Hessian eigenmaps:locally linear embedding techniques for high-dimensional data[J].Proc eedings of the national academy of sciences of USA,2003,100(10):5591—5596.
• [5]ZHANG Z Y,ZHA H Y.Principal manifolds and nonlinear dimension reduction via local tangent space alignment[J].SIAM journal:scientific computing,2005,26(1):313—338.
• [6]李峰,田大庆,王家序.基于有监督增量式局部线性嵌入的故障辨识[J].振动与冲击,2013,32(23):82—88.
• [7]LAW M H C,JAIN A K.Incremental nonlinear dimensionality reduction by manifold learning[J].IEEE transaction on pattern analysis and machine intelligence,2006,28(3):377—391.
• [8]ZHANG Z Y,WANG J,ZHA H Y.Adaptive manifold learning[J].IEEE transaction on pattern analysis and machine intelligence,2012,34(2):253—265.
• [9]侯越先,吴静怡.基于局域主方向重构的适应性非线性维数约简[J].计算机应用,2006,26(4):895—897.
• [10]蒲玲.自适应局部线性降维方法[J].计算机应用与软件,2013,30(4):255—257.
• [11]蒲玲.自适应邻域图的流形学习方法[J].计算机应用与软件,2014,31(2):191—194.
• [12]惠康华,肖柏华,王春恒.基于自适应近邻参数的局部线性嵌入[J].模式识别与人工智能,2010,23(6):842—846.
• [13]文贵华,江丽君,文军.邻域参数动态变化的局部线性嵌入[J].软件学报,2008,19(7):1666—1673.
• [14]MA J J,GONG M G,JIAO L C.Evolutionary clustering algorithm based on mixed measures[J].International journal of intelligent computing and cybernetics,2011,4(4):511—526.
• [15]SOOMIN LEE,ANGELIA N.Distributed random projection algorithm for convex optimization[J].IEEE J Sel Topics Signal Processing,2013,7(2):221—229.
• [16]SANJOY D,ANUPAM G.An elementary proof of the JL lemma:technical report TR-99-006[R].University of California,Berkeley,1990.
• [17]HEGDE C,WAKIN M B,TENENBAUM J.Random projection for manifold learning proofs and analysis:technical report TREE0710[R].Rice University,Houston,2007.
• [18]魏莱,王守觉.基于流形距离的半监督判别分析[J].软件学报,2010,21(10):2445—2453.
• [19]RICHARD G,BARANIUK,MICHAEL B W.Random projections of smooth manifold[J].Foundations of computational mathematics,2009,9(1):51—77.
• [20]DASGUPTA S,GUPTA A.An elementary proof of the Johnson-Lindentrauss lemma:technical report TR-99-006[R].International Computer Science Institute,Berkeley,California,1999.
• [21]FRANKL P,MAEHARA H.The Johnoson-Lindenstrauss lemma and the sphericity of some graphs[J].Journal of combinatorial theory:Ser.B,1988,44:355—362.
• [22]袁暋,程雷,朱然刚.一种新的基于MMC和LSE的监督流形学习算法[J].自动化学报,2013,39(12):2077—2089.
• [23]张光耀,李镇.一个n*n矩阵特征值问题的迹公式及其应用[J].郑州大学学报(理学版),2010,,4(4):18—23.